Prostate carcinogenesis involves inactivation of tumor suppressors and/or activation of oncogenes and is also profoundly influenced by androgens. The mechanism by which androgens affect prostate cancer remains poorly understood. Preliminary studies show that the androgen-response gene Eaf2, also termed U19 (up-regulated gene 19), acts as a potential tumor suppressor in the prostate. Eaf2 down-regulation and allelic loss were detected in >80% of human advanced prostate cancer specimens. Eaf2 knockout (KO) mice developed lymphoma, hepatocellular carcinoma, and poorly differentiated lung adenocarcinoma, demonstrating that Eaf2 is a tumor suppressor in vivo. Although the Eaf2 KO prostate did not develop prostate cancer, it exhibited epithelial hyperplasia and dysplasia (mPIN), suggesting that Eaf2 contributes to the suppression of prostate tumors. Eaf2, along with Eafl, was independently identified and termed ELL- associated factors (Eaf), based on its association with ELL, a fusion partner of MLL in acute myeloid leukemia. Eaf2 and Eaf1 proteins, with 58% identity and 74% similarity, are the only members in the Eaf family. Both Eaf1 and Eaf2 are expressed in prostate epithelial cells, with Eaf1 being constitutive and Eaf2 androgen-inducible. Overexpression of either Eaf1 or Eaf2 induces apoptosis of prostate cancer cells, indicating their functional similarities. These findings led to our hypothesis that tumor suppression requires proper Eaf gene dosage and that the loss of both Eaf family genes can lead to prostate cancer. Four Specific Aims will define the roles of Eaf family genes in prostate carcinogenesis. 1. Determine if an Eaf1 conditional KO in the prostate will cause a similar phenotype to that observed in the Eaf2 KO prostate in the mouse model. 2. Test whether the loss of both Eaf genes causes prostate cancer in the mouse model. We will crossbreed the Eaf2 KO and prostate-specific Eaf1 KO mice to generate Eaf1-Eaf2 double KO mice. 3. Determine the effect of Eaf gene dosage on Eaf-downstream pathways. We will compare the gene expression profiles in the wild-type, Eaf1 KO, Eaf2 KO, and double KO prostate using cDNA microarray. 4. Determine if down-regulation of Eafl and/or Eaf2 in human prostate cancer specimens has prognostic value. We will determine Eafl and/or Eaf2 expression by a semiquantitative immunostaining of human prostate cancer tissue microarray and correlate this with clinical outcomes, such as disease-free survival. The success of this project will define Eaf proteins as tumor suppressors in the prostate, providing a link between androgens and tumor suppressors and a strong foundation for further exploring Eaf pathways for diagnosis, prognosis, prevention, and/or treatment of prostate cancer.